Hi George,
You need to subtract the climatology from your original variable to get the anomaly as is pointed out a bit later in that link.
However, I think your problem is a little bit tricky. It looks like there are large time gaps in your data set. As if you have data for 3 months of the year (1 season?) and then you jump to the next year. I'm not sure that the suggested solution will work for you. I think you might need to create
a temporary monthly time axis from 1979 to 2015 and then map to it. You can then calculate the anomaly and the map back to your original axis.
So maybe something like this (def_monthaxis_days.jnl should come with your Ferret distribution in the contrib directory)
go def_monthaxis_days.jnl noleap 1979 2015 new_time ! Looks like you have a 365 day calendar
let prec_on_monthly = precip[gt=new_time@nrst] ! get values closest
let prec_clim_tmp = prec_on_monthly[gt=month_reg@MOD] !
let prec_clim = prec_clim_tmp[gt=new_time@asn] ! Regrid to monthly axis
let prec_anom_tmp = prec_on_monthly - prec_clim ! still got strange values on months not of interest
let prec_anom = prec_anom_tmp[gt=precip@nrst] ! think this should get you back on the original grid.
plot/X=80/Y=1.0 precip,prec_clim[gt=precip@nrst] ! plot original and climatology
message
plot/X=80/Y=1.0 prec_anom ! Anomaly
This is untested. Somebody else may be able to suggest something if this doesn't work.
Cheers,
Russ
On 21/09/15 15:46, George Otieno wrote:
Dear FerretersI am doing seasonal anaomlaies. When I searched through the archive, When use the climatologcal_axes in example page below
http://ferret.pmel.noaa.gov/Ferret/faq/how-do-i-calculate-climatologies-and-climatological-anomalies
it plots only raw data instead of ANOMALIES.I have used CDO to make seasonal masks. see the snippet below
show data
currently SET data sets:
1> ./GPCP-MAM.nc4 (default)
name title I J K L M N
PRECIP Average Monthly Rate of Precipi 1:144 1:72 ... 1:110 ... ...
yes? show grid
Default grid for DEFINE VARIABLE is ABSTRACT
yes? show grid precip
GRID GEB1
name axis # pts start end
LON LONGITUDE 144mr 1.25E 1.25W
LAT LATITUDE 72 r 88.75S 88.75N
normal Z
TIME TIME 110 i 01-MAR-1979 00:00 01-APR-2015 00:00
normal E
normal F
yes? USE climatological_axes
*** NOTE: regarding /opt/Ferret/V693/go/climatological_axes.cdf ...
*** NOTE: Climatological axes SEASONAL_REG, MONTH_REG, MONTH_IRREG, MONTH_GREGORIAN, MONTH_NOLEAP, MONTH_360_DAY, MONTH_ALL_LEAP defined
yes? CANCEL DATA climatological_axes
yes? LET prec1 = precip[T=01-mar-1979:01-april-2015]
yes? PLOT/X=80/Y=1.0 prec1
yes? LET prec1 = precip[T=01-mar-1979:01-mar-2015]
yes? PLOT/X=80/Y=1.0 prec1
yes? LET prec1_clim = prec1[GT=month_reg@MOD]
yes? PLOT/X=80/Y=1.0/OVERLAY prec1_clim[T=01-mar-1979:01-mar-2015]
I request for any help on how to fix this the probelm
see the plot attached. I to plot only anomalies not raw data.
regards,
George Otieno
IGAD Climate Prediction & Applications Centre (ICPAC)
P.O Box 10304, G.P.O. 00100 Nairobi, KENYA
EMAIL: gotieno@icpac.net
MOBILE: 254 726-902-540
SKYPE: george.otieno204